The applied and theoretical development of statistical methodology is progressing in the areas of biological and epidemiological models, mixture models for describing age changes in distributions of biological markers of morbidity and mortality, multiple comparisons, survival analysis, and the design of experiments, each of which is applicable to longitudinal studies and other studies of aging. The research utilizes various types of statistical models- regression models for studying risk factors' association with outcomes observed in prospective studies, and mixed- effects models for longitudinal data which consider both within- and between-subject variation in analyzing the repeated measurements for all individuals in the study population. Other techniques used to include Bayesian, maximum likelihood and numerical computing methods. The methodology created provides original contributions to experimental testing associated with longitudinal studies, simultaneous comparison of various specified experimental effects, epidemiological study of disease states, survival or failure analyses of longitudinal observations representing growth, physical and mental disability, and other biological and behavioral changes in humans and animals. A major emphasis of this research project is on the development of methods which yield cogent yet easily understood results when applied to data. A new statistical method has been completed for multiple comparisons arising from an experiment involving several treatments or procedures and a control. The new method is an adaptive one which avoids under- or over- conservation which arises in classical comparisonwise or experimentwise testing rules.